The aircraft wreckage was examined to the extent possible, given the massive destruction and loss by fire. No evidence of system malfunction or pre-existing defect was found. All flight control surfaces were accounted for at the accident site and no sign of in-flight detachment or malfunction was identified. The sightings of the aircraft seconds before impact confirm that it was intact, without sign of fire, with all engines operating, and with the hydraulic and electrical systems functioning. Further, the fact that the pilot flying released two loads before impact signifies that he was conscious and aware of the situation they were in at that time. Since there was no communication or witness accounts to the contrary, nor was the emergency dump system employed, it would be reasonable to conclude that, in this relatively simple and benign flight task, the pilots had not experienced an aircraft flight control malfunction. Additionally, the low rate of climb demonstrated by the aircraft - in the absence of mechanical reason - suggests that the pilots were not aware of their proximity to the rising terrain. A discussion of these two issues follows. Given the flight time and the distance to the impact point, the permutations of the radius of turn for various airspeeds show that the Electra would have described an arc over the ground dictated by two variables - airspeed and bank angle - with the wind as a constant. Accordingly, the most likely airspeed-bank combinations range from 135to 150knots and 40to 45degrees, neither of which is unreasonable for this type of operation. Such a flight path is consistent with purposely controlled flight. Values outside these finite ranges were found to produce great divergence from the last minute of flight path and were discounted. The exit route down the valley toward Moyie Lake was always available to the pilots in the event of a mechanical malfunction that required manoeuvring room over flat terrain. Accordingly, since the aircraft did not deviate from its arcing path to the left, it is unlikely that the pilots were dealing with a situation preventing climb performance or with a flight control malfunction. Furthermore, the fact that the aircraft was seen flying in a conventional attitude until shortly before impact suggests that the pilots were in control of the aircraft. Examinations of the engines and propellers consistently show that the engines were delivering similar and high power at impact. The reasonable conclusion is, therefore, that the engines were capable of delivering rated power before the impact. In concert with the airspeed and angle of bank calculations, the results of the rate of climb assessments reveal that the Electra could have attained climb rates in the order of 1000fpm at any stage during the left turn. Accordingly, with consideration of the known flight path and altitude profile, no mechanical reason, aerodynamic effect, or performance factor was found that would have prevented the aircraft from attaining sufficient altitude to avoid the terrain. Without the likelihood of either aircraft mechanical defect or flight performance inadequacy, the explanation as to why the aircraft did not climb to avoid the terrain becomes one of human performance. With an aircraft capable of controlled, climbing flight, it would be reasonable to conclude that the pilots flew the vertical profile after the drop on the target fire and gently climbed from 3700feet to at least 3900feet. It could not be determined if the aircraft climbed more than the observed 200feet difference, only to lose it again. It is also possible that, after initiating a satisfactory rate of climb for the seemingly benign task to climb to 4500feet, the pilots became distracted in the cockpit and unintentionally allowed the climb to deteriorate. Because of the lack of direct knowledge of the cockpit circumstances leading to the accident, the reason for the pilots not climbing their aircraft cannot be identified with any certainty. Nonetheless, several factors exist that collectively lead to a possible explanation. Documented research10 shows that mountainous terrain in daylight, especially at midday with little shadow outline, lacks effective visual definition and is particularly difficult for pilots to assess for slope, proximity, and rate of closure. In such conditions, visual illusion and depth misperception is quite likely. In this accident, the characteristics of the terrain were similarly difficult to assess, and it is most likely that the Electra pilots were deceived by the apparent gently sloping nature of the surrounding terrain and did not detect the ridgeline that crossed their flight path. To them, the ridgeline and the protruding land mass would have been difficult to discern clearly until the aircraft was so close that timely and effective evasive action was impossible. Compounding their difficulty in detecting the obstacles ahead of them were the limitations to unimpeded vision resulting from the left bank in the turn toward the rising terrain, which reduced their perspective and field of vision, and the physical dimensions of the cockpit windshields. It can be said with some degree of certainty that the pilots were unaware that they were on a collision path with the terrain until the very last seconds before impact, otherwise they would have altered course and flown the simple, low-level emergency route down the valley toward Moyie Lake. Low-level, fire-management flight operations continually challenge the situational awareness skills of pilots and require assiduous attention to the terrain, aircraft performance, and effective crew resource management. Vision limitations resulting from cockpit design and layout characteristics can further reduce a pilot's opportunity to detect obstacles in the flight path ahead.Analysis The aircraft wreckage was examined to the extent possible, given the massive destruction and loss by fire. No evidence of system malfunction or pre-existing defect was found. All flight control surfaces were accounted for at the accident site and no sign of in-flight detachment or malfunction was identified. The sightings of the aircraft seconds before impact confirm that it was intact, without sign of fire, with all engines operating, and with the hydraulic and electrical systems functioning. Further, the fact that the pilot flying released two loads before impact signifies that he was conscious and aware of the situation they were in at that time. Since there was no communication or witness accounts to the contrary, nor was the emergency dump system employed, it would be reasonable to conclude that, in this relatively simple and benign flight task, the pilots had not experienced an aircraft flight control malfunction. Additionally, the low rate of climb demonstrated by the aircraft - in the absence of mechanical reason - suggests that the pilots were not aware of their proximity to the rising terrain. A discussion of these two issues follows. Given the flight time and the distance to the impact point, the permutations of the radius of turn for various airspeeds show that the Electra would have described an arc over the ground dictated by two variables - airspeed and bank angle - with the wind as a constant. Accordingly, the most likely airspeed-bank combinations range from 135to 150knots and 40to 45degrees, neither of which is unreasonable for this type of operation. Such a flight path is consistent with purposely controlled flight. Values outside these finite ranges were found to produce great divergence from the last minute of flight path and were discounted. The exit route down the valley toward Moyie Lake was always available to the pilots in the event of a mechanical malfunction that required manoeuvring room over flat terrain. Accordingly, since the aircraft did not deviate from its arcing path to the left, it is unlikely that the pilots were dealing with a situation preventing climb performance or with a flight control malfunction. Furthermore, the fact that the aircraft was seen flying in a conventional attitude until shortly before impact suggests that the pilots were in control of the aircraft. Examinations of the engines and propellers consistently show that the engines were delivering similar and high power at impact. The reasonable conclusion is, therefore, that the engines were capable of delivering rated power before the impact. In concert with the airspeed and angle of bank calculations, the results of the rate of climb assessments reveal that the Electra could have attained climb rates in the order of 1000fpm at any stage during the left turn. Accordingly, with consideration of the known flight path and altitude profile, no mechanical reason, aerodynamic effect, or performance factor was found that would have prevented the aircraft from attaining sufficient altitude to avoid the terrain. Without the likelihood of either aircraft mechanical defect or flight performance inadequacy, the explanation as to why the aircraft did not climb to avoid the terrain becomes one of human performance. With an aircraft capable of controlled, climbing flight, it would be reasonable to conclude that the pilots flew the vertical profile after the drop on the target fire and gently climbed from 3700feet to at least 3900feet. It could not be determined if the aircraft climbed more than the observed 200feet difference, only to lose it again. It is also possible that, after initiating a satisfactory rate of climb for the seemingly benign task to climb to 4500feet, the pilots became distracted in the cockpit and unintentionally allowed the climb to deteriorate. Because of the lack of direct knowledge of the cockpit circumstances leading to the accident, the reason for the pilots not climbing their aircraft cannot be identified with any certainty. Nonetheless, several factors exist that collectively lead to a possible explanation. Documented research10 shows that mountainous terrain in daylight, especially at midday with little shadow outline, lacks effective visual definition and is particularly difficult for pilots to assess for slope, proximity, and rate of closure. In such conditions, visual illusion and depth misperception is quite likely. In this accident, the characteristics of the terrain were similarly difficult to assess, and it is most likely that the Electra pilots were deceived by the apparent gently sloping nature of the surrounding terrain and did not detect the ridgeline that crossed their flight path. To them, the ridgeline and the protruding land mass would have been difficult to discern clearly until the aircraft was so close that timely and effective evasive action was impossible. Compounding their difficulty in detecting the obstacles ahead of them were the limitations to unimpeded vision resulting from the left bank in the turn toward the rising terrain, which reduced their perspective and field of vision, and the physical dimensions of the cockpit windshields. It can be said with some degree of certainty that the pilots were unaware that they were on a collision path with the terrain until the very last seconds before impact, otherwise they would have altered course and flown the simple, low-level emergency route down the valley toward Moyie Lake. Low-level, fire-management flight operations continually challenge the situational awareness skills of pilots and require assiduous attention to the terrain, aircraft performance, and effective crew resource management. Vision limitations resulting from cockpit design and layout characteristics can further reduce a pilot's opportunity to detect obstacles in the flight path ahead. For undetermined reasons, the Electra did not climb sufficiently to avoid striking the rising terrain. Given the flight path and the rate of climb chosen, a collision with the terrain was unavoidable. The characteristics of the terrain were deceptive, making it difficult for the pilots to perceive their proximity and rate of closure to the rising ground in sufficient time to avoid it.Findings as to Causes and Contributing Factors For undetermined reasons, the Electra did not climb sufficiently to avoid striking the rising terrain. Given the flight path and the rate of climb chosen, a collision with the terrain was unavoidable. The characteristics of the terrain were deceptive, making it difficult for the pilots to perceive their proximity and rate of closure to the rising ground in sufficient time to avoid it. Performance calculations show that the Electra - in the absence of limiting mechanical malfunction - could have climbed at a rate that would have allowed the aircraft to avoid the terrain. Although a functional cockpit voice recorder was installed in the aircraft, it was not required by regulation and it was not used; as a result, vital clues that could have shed light on the circumstances of this accident were not available. The emergency locator transmitter could not transmit a signal as a result of severe impact forces that exceeded the design criteria.Other Findings Performance calculations show that the Electra - in the absence of limiting mechanical malfunction - could have climbed at a rate that would have allowed the aircraft to avoid the terrain. Although a functional cockpit voice recorder was installed in the aircraft, it was not required by regulation and it was not used; as a result, vital clues that could have shed light on the circumstances of this accident were not available. The emergency locator transmitter could not transmit a signal as a result of severe impact forces that exceeded the design criteria. Safety Action Taken British Columbia Forest Service The Aviation Management Division of British Columbia Forest Service (BCFS) participated in the TSB investigation and conducted an internal examination of the circumstances to identify and correct any BCFS-related flight safety factors. Its report found that the flight operations practices were in accordance with the BCFS Firebombing Procedures Manual, with the exception that the Electra turned left before attaining sufficient altitude to avoid the rising terrain. The report based on the internal examination by BFCS recommended that the BCFS firebombing training and procedures be reinforced so that, if tanker pilots choose to take an exit or line that is different from that demonstrated by the bird dog aircraft, they must advise the bird dog crew to allow them the option of re-checking the proposed (new) route for hazards. It was further recognized that bird dog and tanker aircraft may differ substantially in flight performance, and that, accordingly, the Spring2004 operational air attack training sessions should focus on safety procedures and potential risks in the low-level firebombing environment. BCFS aviation occurrence information reveals several instances where tanker pilots have been reticent to dump the retardant load and crashed as a result. On the other hand, there are data to show that releasing the load has prevented an occurrence. Although in this accident the pilots using the emergency dump mechanism would have had little effect on the outcome, the BCFS also recommended that pre-season training and operational practices conducted by their aircraft operators include the practical use of the aircraft emergency dump systems. Crew Resource Management (CRM) training was introduced in the BCFS air tanker program in1993, and the success of this program initiative is evident in that no serious injury had occurred since. Further, BCFS has developed an effective pilot/crew decision-making (PDM) program tailored to fire and forestry aviation operations. BCFS believes that such workshops and training are essential for safe fire-management aviation operations in British Columbia and recommended that the air tanker operators continue to provide initial and refresher CRM and PDM training to their pilots. BCFS's Aviation Management Division actioned the recommendations outlined above in the following manner: At the April/May2004 operational air attack training sessions, the firebombing procedures were reinforced, as well as the priorities of runs and exits, specifically in situations where tanker pilots choose to take a different exit or line. They must advise the bird dog crew to allow them the option of re-checking the proposed route for hazards. The pre-season training and operational practices conducted by contracted pilots include the practical use of the aircraft emergency dump systems. The BCFS Air Tanker Program and the air tanker operators are actively monitoring the delivery of initial and refresher PDM and CRM training to all pilots and air attack staff in the British Columbia program. Air Spray As a result of this accident, and commencing with its 2004annual pilot training course, AirSpray has placed additional emphasis on human factors and emergency manoeuvring in mountainous areas. Particular attention has been given to the deceptive nature of mountainous terrain at high sun angles, and the deceptive illusionary nature of mountain flying continues to be stressed in its training programs.